Korvus

Unified RAG Pipeline

Korvus is a search SDK that enables users to execute the entire Retrieval Augmented Generation (RAG) pipeline with a single database query. This allows for streamlined and efficient data retrieval and processing.

Postgres Integration

Korvus is built on top of PostgreSQL, a robust and popular open-source relational database system. This integration offers advantages such as scalability, reliability, and data integrity.

Multi-Language Support

Korvus provides bindings for various programming languages, including Python, JavaScript, Rust, and C. This allows developers to utilize the SDK in their preferred language environments.

Vector Similarity Search

The SDK supports vector similarity search, a powerful technique for finding relevant information based on semantic similarity rather than exact keyword matching. This feature is particularly useful for natural language processing (NLP) tasks.

Embedding Integration

Korvus seamlessly integrates with embedding models, which map text to numerical vectors. This integration allows the SDK to perform efficient and accurate semantic search within the database.

Community and Resources

Korvus has an active community on Discord and Twitter. These platforms provide a space for users to discuss the SDK, share insights, and collaborate on projects.

• Application Developers: Implement a RAG pipeline with the Korvus SDK, utilizing its Python and JavaScript bindings.
• Data Scientists: Build scalable, high-performance search applications leveraging Korvus's single query RAG capabilities on Postgres.
• Enterprise Architects: Simplify complex architectures by replacing service oriented approaches with Korvus's unified Postgres-native RAG pipeline.
• Machine Learning Engineers: Customize and extend Korvus's SQL operations for advanced RAG functionality and improved developer experience.
• Open Source Contributors: Contribute to the Korvus project by enhancing multi-language support and improving existing features.

What is Korvus?

Korus is a search SDK that unifies the entire RAG pipeline in a single database query. It is built on top of Postgres with bindings for Python, JavaScript, Rust and C.

How do I use Korvus?

You can use Korvus by installing the library and using the provided API. Korvus is designed to be easy to use, with a simple API that allows you to quickly get started with RAG.

What are the benefits of using Korvus?

Korus offers a number of benefits over other RAG solutions, including:

• Unified pipeline: Korus unifies the entire RAG pipeline in a single database query. This makes it much easier to use and manage.
• High performance: Korus is built on top of Postgres, which is a high-performance database. This ensures that your RAG queries will be executed quickly.
• Flexibility: Korus supports multiple languages, including Python, JavaScript, Rust and C. You can use the language that best suits your needs.

How does Korvus compare to other RAG solutions?

Korus is a unique RAG solution that offers a number of advantages over other solutions. For example, Korus is the only RAG solution that unifies the entire RAG pipeline in a single database query. This makes it much easier to use and manage. Korus is also built on top of Postgres, which is a high-performance database. This ensures that your RAG queries will be executed quickly.

Where can I learn more about Korvus?

You can learn more about Korvus by visiting the Korus website. You can also join the Korus Discord server or follow the Korus Twitter account.

Korvus is a search SDK designed to unify the RAG pipeline, using a single database query. It leverages Postgres, offering bindings for Python, JavaScript, Rust and C, to deliver efficient search capabilities.

• Ensure you have a Postgres database with pgml and pgvector installed, either self-hosted or via a managed service like PostgresML Cloud.
• Install the Korvus package using pip: pip install korvus. This provides the necessary Python bindings for interacting with Korvus.
• Set the KORVUS_DATABASE_URL environment variable with your database connection string to allow Korvus to connect.
• Initialize a Collection and Pipeline, defining the data source and processing steps for your RAG operations, including splitting and semantic search.
• Insert or update documents using collection.upsert_documents(), ensuring your data is available for retrieval and augmented generation.
• Perform Retrieval-Augmented Generation (RAG) using collection.rag() to retrieve relevant context and generate responses based on your data.
• Review the results. Korvus combines context retrieval and text generation in a single query, simplifying RAG and improving performance.
• Customize SQL operations for advanced control, taking advantage of PostgreSQL's query optimization capabilities to improve performance and tailor results.